You are on page 1of 66

Graduate Theses and Dissertations Graduate College

2014

Fabrication and angle compensation analysis of


skew rolling mill
Jie Wang
Iowa State University

Follow this and additional works at: http://lib.dr.iastate.edu/etd


Part of the Engineering Mechanics Commons, Mechanical Engineering Commons, and the
Mechanics of Materials Commons

Recommended Citation
Wang, Jie, "Fabrication and angle compensation analysis of skew rolling mill" (2014). Graduate Theses and Dissertations. 13927.
http://lib.dr.iastate.edu/etd/13927

This Thesis is brought to you for free and open access by the Graduate College at Iowa State University Digital Repository. It has been accepted for
inclusion in Graduate Theses and Dissertations by an authorized administrator of Iowa State University Digital Repository. For more information,
please contact digirep@iastate.edu.
Fabrication and angle compensation analysis of skew rolling mill

by

Jie Wang

A thesis submitted to the graduate faculty

in partial fulfillment of the requirements for the degree of

MASTER OF SCIENCE

Major: Mechanical Engineering

Program of Study Committee:


Gap-Yong Kim, Co-major Professor
Abhijit Chandra, Co-major Professor
Kaitlin Bratlie

Iowa State University

Ames, Iowa

2014

Copyright © Jie Wang, 2014. All rights reserved.


ii

TABLE OF CONTENTS

LIST OF FIGURES ................................................................................................................. iv

LIST OF TABLES ................................................................................................................... vi

NOMENCLATURE ............................................................................................................... vii

ABSTRACT ............................................................................................................................. ix

CHAPTER 1. INTRODUCTION ............................................................................................. 1


1.1. Motivation ...................................................................................................................... 1
1.2. Research framework and objectives ............................................................................... 2
1.2.1. General design of a skew rolling mill ...................................................................... 2
1.2.2. Experimental validation of force analysis ............................................................... 3
1.2.3. Study of the angle compensation ............................................................................. 3
1.3. Dissertation organization................................................................................................ 3

CHAPTER 2. LITERATURE SURVEY.................................................................................. 5


2.1. Rolling ............................................................................................................................ 5
2.2. Friction in rolling ........................................................................................................... 7
2.3. Skew rolling ................................................................................................................... 8

CHAPTER 3. GENERAL DESIGN OF A SKEW ROLLING MILL ................................... 11


3.1. General design .............................................................................................................. 11
3.2. Design of the power source with speed adjustment ..................................................... 11
3.3. Design to change reduction ratios ................................................................................ 13
3.4. Design to realize the skewness function of the setup ................................................... 14
3.5. Design to acquire forces ............................................................................................... 15
3.5.1. Dynamometer ........................................................................................................ 15
3.5.2. Data acquisition interface ...................................................................................... 16
3.5.3. Dynamometer calibration ...................................................................................... 18

CHAPTER 4. FORCE ANALYSIS AND GEOMETRY MODELS IN THE ROLLING


PROCESS ............................................................................................................................... 19
4.1. Force analysis and geometry model of parallel rolling ................................................ 19
4.2. Force analysis and geometry model of skew rolling .................................................... 22
4.3. Experimental validation of force analysis .................................................................... 28
4.3.1. Experimental data, calculations and simulations of parallel rolling ...................... 28
4.3.2. Experimental data, calculations and simulations of skew rolling process ............ 35
iii

CHAPTER 5. EXPERIMENTAL STUDY ON THE ANGLE COMPENSATION .............. 42


5.1. Experimental procedure and results ............................................................................. 42
5.2. Discussion .................................................................................................................... 49

CHAPTER 6. CONCLUSIONS ............................................................................................. 52

CHAPTER 7. BIBLIOGRAPHY............................................................................................ 54
iv

LIST OF FIGURES

Fig. 2.1 Skew rolling mill with groves ..................................................................................... 9

Fig. 3.1 Scheme for the motor and the bottom roller of the skew rolling setup ..................... 12

Fig. 3.2 Detailed view of the Toshiba VFD used to adjust the RPM ..................................... 13

Fig. 3.3 Shims used in the setup to adjust gap between two rollers ....................................... 14

Fig. 3.4 Solidworks drawing for the upper plate to achieve required angle ........................... 15

Fig. 3.5 Force measurement setup flowchart in the rolling mill ............................................ 16

Fig. 3.6 LabVIEW interface front panel. The white, red, greed curves stand for
X, Y, and Z forces ................................................................................................................... 17

Fig. 3.7 Block diagram behind front panel of the LabVIEW program ................................... 17

Fig. 3.8 Dynamometer shows the sensitivities and amplification factors of three axes ......... 18

Fig. 4.1 Deformation scheme in parallel rolling process ........................................................ 20

Fig. 4.2 Schematic drawing of Sp ............................................................................................ 20

Fig. 4.3 Forces analysis on bottom roller in parallel rolling process ...................................... 21

Fig. 4.4 Specimen in between the two rollers, the red rectangle is the maximum area
it can deform, the arc shaped area is an approximate area assumed ....................................... 23

Fig. 4.5 Specimen in contact with the bottom roller surface. Contact curve is
represented as the ellipse dot line ........................................................................................... 24

Fig. 4.6 Force analysis on bottom roller under skew rolling process .................................... 26

Fig. 4.7 Rolling model created in Abaqus .............................................................................. 28

Fig. 4.8 Measured Y axis force under parallel rolling with different reductions................... 30

Fig. 4.9 Measured Z axis force under parallel rolling with different reductions .................... 30

Fig. 4.10 Compressive stress strain curve for HDPE, green curve in the graph is applied .... 32

Fig. 4.11 Compressive stress strain curve for PETG, 0.5/s plane in the graph is applied ...... 32
v

Fig. 4.12 Comparison among the measured forces, the calculated forces and
simulated forces in Y axis in parallel rolling .......................................................................... 34

Fig. 4.13 Comparison among the measured forces, the calculated forces and
simulated forces in Z axis in parallel rolling .......................................................................... 35

Fig. 4.14 Measured force in Y axis of PETG under different angles .................................... 36

Fig. 4.15 Measured force in Y axis of HDPE under different angles .................................... 37

Fig. 4.16 Measured force in Z axis of PETG under different angles ..................................... 37

Fig. 4.17 Measured force in Z axis of HDPE under different angles .................................... 38

Fig. 4.18 Comparison between the measured forces and the calculated forces in Y axis
in skew rolling for PETG and HDPE at three different angles ............................................... 39

Fig. 4.19 Comparison between the measured forces and the calculated forces in Z axis
in skew rolling for PETG and HDPE at three different angles ............................................... 40

Fig. 5.1 X axis force of HDPE at an angle of 6 ..................................................................... 43

Fig. 5.2 Z axis force of HDPE at an angle of 3 ..................................................................... 43

Fig. 5.3 X axis force of PETG at an angle of 3 .................................................................... 44

Fig. 5.4 Z axis force of PETG at an angle of 8 ..................................................................... 44

Fig. 5.5 X axis forces of HDPE at different angles ................................................................ 46

Fig. 5.6 Z axis forces of HDPE at different angles ................................................................. 46

Fig. 5.7 X axis forces of PETG at different angles ................................................................. 47

Fig. 5.8 Z axis forces of PETG at different angles ................................................................. 47

Fig. 5.9 Comparison between simulation and experiment data of Fxd/ Fzd of HDPE ............. 48

Fig. 5.10 Comparison between simulation and experiment data of Fxd/ Fzd of PETG............ 49
vi

LIST OF TABLES

Table 3.1 Specifications of the dynamometer employed in the setup .................................... 16

Table 4.1 Experimental settings of parallel rolling for HDPE and PETG .............................. 29

Table 4.2 Comparison between the measured forces and the calculated forces in parallel
rolling ...................................................................................................................................... 33

Table 4.3 Experiment setting of skew rolling process with PETG and HDPE....................... 36

Table 4.4 Comparison between the measured forces and the calculated Y axis and Z axis
forces for PETG and HDPE at three different angles ............................................................. 39

Table 5.1 Experiment setting of two specimens ..................................................................... 42

Table 5.2 Experiment X and Z axes forces of HDPE and PETG at eight different angles .... 45
vii

NOMENCLATURE

h, thickness

hmax, maximum thickness

h0, specimen initial thickness

hf, specimen final thickness

w, specimen width

R, roller radius

x, unknown parameter in the equation

Sp, deformed area under parallel rolling

Ss, deformed area under skew rolling

μ, coefficient of friction

, crossed angle between the upper roller and bottom roller

P, pressure on the specimen

Ra, area ratio

Fxp, calculated force in X axis direction in the center under parallel rolling

Fyp, calculated force in Y axis direction in the center under parallel rolling

Fzp, calculated force in Z axis direction in the center under parallel rolling

Fxd, recorded force in Z axis direction at one end of the bottom roller by dynamometer

Fyd, recorded force in Y axis direction at one end of the bottom roller by dynamometer

Fzd, recorded force in Z axis direction at one end of the bottom roller by dynamometer

Fxs, calculated force in X axis direction in the center under skew rolling

Fys, calculated force in Y axis direction in the center under skew rolling

Fzs, calculated force in Z axis direction in the center under skew rolling
viii

ACKNOWLEDGEMENTS

First of all, I would like to thank Dr. Gap-Yong Kim and Dr. Abhijit Chandra, my

two advisors, for their guidance, patience, and especially, for offering me the opportunity to

work on these projects. Second, I would like to thank my committee member, Dr. Kaitlin

Bratlie, for her efforts and contributions to this work. I would also like to thank Wyman

Martinek, Craig Severson, and Jim Dautremont in the department who provided significant

help for this work. I want to offer my appreciation to my colleagues and my friends as well,

especially Mina Bastwros. The work by Jack B. Lorch, Letian Yang, and Jianan Wu is

greatly acknowledged.
ix

ABSTRACT

The rolling process is an excellent method for processing new high performance

polymers. Skew rolling, as a highly accurate and highly efficient rolling technology, has been

commonly used to enhance mechanical properties and productivity. However, the complexity

of the mechanics in the skew rolling process of polymers is a significant challenge for

manufacturers and researchers. In this work, models for calculating the deformed area under

parallel rolling and skew rolling have been developed. Based on the models, interactive force

models in the rolling process have also been developed. Finite element analysis was used to

simulate the rolling process. Experimental results were compared with the force model and

simulation results. A study on angle compensation based on the influences of angle and

friction conditions on force ratio Fxd/Fzd in the skew rolling process was carried out. The

results showed the accuracy of models and equations, and the work offered the possibility of

angle compensation for the two polymers employed in this study.


1

CHAPTER 1. INTRODUCTION

1.1. Motivation

New rolling mills, such as the continuous variable crown (CVC) and pair cross (PC)

[1-3], as well as work rolls crossing and shifting (RCS) [4] mills have been introduced and

developed to improve the strip shape and profile. The PC mill was first used commercially at

Nippon Steel’s Kimitsu Works in 1991 [1], and it has been commonly used in hot rolling

mills for plating and stripping as a finishing mill. In the development of highly accurate and

highly efficient rolling technology, the PC mill significantly improved the capability of the

shape and crowns control, reduced the edge drop, and also enhanced mechanical properties

and productivity [1]. This system has been applied not only to hot strip mills [5, 6] but also to

plate mills and cold strip mills and the areas of application are still increasing. Research has

focused on analyzing the rolling load or rolling torque generated by pair cross mills [7, 8],

and on the strip profile control capability of the roll crossing mill[9], and so forth.

Polymers have increasingly been used to replace metallic parts in a wide range of

applications over the past few decades. The rolling process is one of the best methods for

processing new high performance polymers. The process allows for fabrication of more

highly valued products than products produced by simple injection [10] and it can produce

components more accurately and effectively [11]. The rolling process has been used for

many other materials such as polypropylene (PP), polyethylene and polyoxymethylene due to

the possibility of achieving high molecular orientation [12].

However, only a small amount of research literature was found on the skew rolling

process of polymers. The complexity of the mechanics of deformation is a significant


2

challenge to the skew rolling process of polymers. The motivation of this research is to study

the influence of crossed angle and friction conditions on the mechanics of polymers in the

skew rolling process. In this work, a geometric model of deformed specimens will be built

and forces will be analyzed. Methods to control the forces in the skew rolling process will be

proposed based on the investigation of forces.

1.2. Research framework and objectives

The objective of this research is to study the influence of crossed angle and friction

conditions on the force ratio Fxd/Fzd in order to offer methods to control axial force. A

simplified force model in parallel rolling was applied in order to compare the measured

forces via dynamometer in order to determine the friction conditions for the two different

materials. A force model in skew rolling was developed based on the friction conditions in

the parallel rolling. Meanwhile, measured data were utilized to validate the models in skew

rolling. With validated models, the influence of crossed angles and friction conditions on the

force ratio Fxd/Fzd was analyzed.

1.2.1. General design of a skew rolling mill

This section focuses on the design and construction of the skew rolling mill to

produce a skew rolling system and to measure forces in three axes via a dynamometer, as

well as to include the basic functions of traditional rolling mills. This section also describes

the calibration process for the dynamometer in detail.


3

1.2.2. Experimental validation of force analysis

Forces in parallel rolling and skew rolling are fundamental characteristics in the

rolling process. Forces characteristics in the rolling process affect the rolling stability,

potentially causing concern for practical applications. This section simplifies and analyzes

force models of parallel rolling and skew rolling. Forces were measured and experimental

outcomes were compared to those models to provide validation.

1.2.3. Study of the angle compensation

Based on the validated models in skew rolling, the influence of a) rolling process

parameters, b) the crossed angle, and c) friction conditions, on the force ratio Fxd/ Fzd were

analyzed. This section captured the angles and friction conditions on the force ratio in order

to demonstrate the feasibility of compensating angles in order to minimize the axial forces on

bearings.

1.3. Dissertation organization

The remainder of this dissertation is divided into five chapters. In Chapter 2, a

literature survey with regard to rolls, skew rolling, and friction in the rolling process is

conducted. In Chapter 3, details of the skew rolling mill are illustrated. Design concepts to

realize required functions are described in detail. In Chapter 4, force models of parallel

rolling and skew rolling were built and compared with experimental behaviors. In Chapter 5,

the influence of angles and friction conditions on the force ratio is studied and the angle
4

compensation is analyzed based on the results. A summary and conclusion are presented in

Chapter 6.
5

CHAPTER 2. LITERATURE SURVEY

2.1. Rolling

Rolling is a process of reducing the thickness or changing the cross section of long

workpieces through compressive forces applied by a set of rolls; thus the process is similar to

rolling dough with a rolling pin to reduce the dough’s thickness. Rolling is widely used due

to its high rate of production and the accurate control of final product [13]. Rolling is used to

produce slabs, sheets, strips, and foils with a dense attractive surface finish and strengthened

mechanical properties. Rolling processes can be divided into hot rolling process and cold

rolling process, based on the rolling temperature.

Hot rolling is a metalworking process that occurs above the recrystallization

temperature of the material. After the grains deform during the rolling process, they

recrystallize and prevent the metal from hardening while being worked. Hot rolled metals

have little directionality in general in terms of their mechanical properties; their residual

stresses are induced by deformation. But in certain instances non-metallic inclusions would

impart a certain amount of directionality, and workpieces less than 20 mm thick often have

some directional properties. Also, a lot of residual stresses to those shapes that have a non-

uniform cross-section are induced by non-uniform cooling. While the finished product is of

good quality, the surface is covered in mill scale, which is the anoxide that forms at high

temperatures. It is usually removed through pickling or by a smooth clean surface process,

which reveals a smooth surface [14]. Dimensional tolerances are usually two to five percent

of the overall dimension [14].


6

Cold rolling is a metalworking process that occurs when the metal’s temperature is

below its recrystallization temperature (usually at room temperature); this process increases

the strength by strain hardening. This process also improves the surface’s finish. Products

commonly produced by cold rolling include sheets, bars, strips, and rods; these products are

usually smaller than products that are produced by hot rolling. Due to the smaller size of the

workpieces and their greater strength compared to hot rolled stock, four-high or cluster mills

are used [14]. It is worth noting that in a single pass, cold rolling cannot reduce the thickness

of a workpiece as much as hot rolling can.

Other shapes can be cold-rolled if the cross-section is relatively uniform and the

transverse dimension is relatively small. Cold rolling shapes requires a series of shaping

operations, usually along the lines of sizing, breakdown, roughing, semi-roughing, semi-

finishing, and finishing.

Although about 90% of all metals produced by the metalworking processes are

produced through a rolling process [15], the rolling process has also been used for many

other materials including polypropylene (PP), polyethylene and polyoxymethylene due to the

possibility of achieving a high level of molecular orientation [12]. Rolling is an excellent

method for processing new high performance polymers. It can be used to manufacture

products that are more highly valued than products produced by simple injection [10]. The

rotary forming process has attracted significant interest in recent years because this process

can produce components more accurately and effectively [11]. Polymers have increasingly

been used to replace metallic parts in a wide range of applications over the past few decades.

Polymers have been used in rolling processes to produce materials with various performances

characteristics [16].
7

Ring rolling is a special forging process that has demonstrated the feasibility of cold

processing of certain polymers [11]. Research on polymers subjected to cold rolling has been

done extensively: plastic deformation and morphology change of crystalline polymer

materials by rolling processes were investigated by J. Qiu et al. [10, 12]. Amorphous

transitions in glassy polymers subjected to cold rolling was studied by D. Cangialosi et al.

[17]. Elastic-recovery properties and dimensional changes followed by the cold rolling of

rods of thermoplastics of various kinds were studied [18]. Fatigue crack growth in

polycarbonate (PC) and polyvinylchloride (PVC) have been studied using anisotropic sheets

oriented by cold rolling [19]. Deformation of semicrystalline polymerics subjected to cold

rolling were studied by T. Asano and Y. Fujiwara [20]. The tensile behavior of high density

polyethylene subjected to cold rolling was studied by R. M. Caddell et al. [21]. The effects of

cold rolling on crazing of polycarbonate was studied by G. O. Shonaike and P. E. Reed [22].

2.2. Friction in rolling

Friction is due to the interactions between the opposite asperities of the two sliding

surfaces. From the interactions, friction is divided into two categories: adhesion and

deformation [23]. Adhesion is the characteristic that some asperities of the contact surfaces

will adhere when there is load in the two objects. There would be a junction between the two

objects that will impede the motion of the two objects. Friction is generated as a result of the

relative motion of one object over another object. The second kind is deformation in the

macroscopic interaction. The deformation will cause the harder surface to plough grooves in

the softer surface.


8

K. L. Johnson studied the friction in rolling contact in detail [24]. In Johnson’s

opinion, all the motions between two contact bodies can be divided into sliding, rolling, and

spinning. The resultant forces can always be resolved into a normal force, which acts in the

direction of the normal on the initial contact point of the two contact bodies, and a tangential

force, which acts on the tangential plane. If the normal force is zero, the motion is known as

free rolling, while the other is called tractive rolling. In free rolling, the contact stress and

deformation can be calculated using Hertz theory [25].

F. W. Carter [26], H. Poritsky [27], K. L. Johnson [24] and R. D. Arnell [28] have

studied the tractive rolling of elastic cylinders in detail. In the contact zone of tractive rolling,

there is a central “stick” area and two outer “slip” areas. This is possible because of the

deformable nature of materials. Thus, if the normal force is constant while the tangential

force is increased from zero, microslip occurs at the edges of the contact zone and spreads

inward as the tangential force increases; when the tangential force reaches the limit value, the

two zones of slip meet in the center and gross sliding begins.

Rolling resistance also comes from the adhesive and deformation losses even though

the magnitude is smaller comparatively. Deformation loss (as opposed to adhesive loss) is the

major loss in the resistance of free rolling. In tractive rolling, deformation loss is also the

dominant factor of resistance.

2.3. Skew rolling

The skew rolling process, evolved from the transverse rolling process, is a unique

rolling process that produces near-net-shape cylindrical or annular metal components and the
9

process has the ability to contour the outside diameter. The name “skew rolling” reveals one

of the main characteristics of this process: that the axes of the rolls are arranged obliquely to

each other with a small angle [29]. Fig. 2.1 shows an image of the skew rolling with grooved

rollers. Worth mentioning here is that in our setup, the cylindrical rollers are plain surfaces

without grooves.

Fig. 2.1 Skew rolling mill with groves

Skew rolling is a new technology for producing various shafts. Compared to

traditional casting, forging, and machining, this process has many remarkable advantages in

productivity, material consumption, overall mechanical properties and service life [30].

Various kinds of products ranging from balls, rollers, stepped shafts, to anchor rods with

threads are being manufactured by skew rolling [30].

However, the complexity of the forming process is a notable challenge for all

manufacturers when developing skew rolling processes. Theories that explaining the

mechanics and tool workpiece interactions are mainly based on simplified assumptions and

empirical results. Numerous variables in the process are another challenge the manufacturers

face.
10

To tackle those challenges, the most powerful and most widely employed tool is the

finite element method. The finite element method is widely used to characterize the

workpiece material stress, strain, and deformation behavior [31]. Another method is to

analyze the skew rolling method through mathematical models derived and simulated based

on envelope theories and assumptions [32]. To design the parameters in the skew rolling

process, mathematical algorithms are also widely employed to optimize the design process

[33].
11

CHAPTER 3. GENERAL DESIGN OF A SKEW ROLLING MILL

3.1. General design

According to the aim of the experiments for cold skew rolling, the rolling mill should

have the following abilities:

1. The ability to roll the specimens at an adjustable speed;

2. The ability to change the gap between two rollers in the range of 1mm-5mm;

3. The ability to skew the rollers along a range of angles from -15-15;

4. The ability to measure the forces along three axes

3.2. Design of the power source with speed adjustment

To provide power for the rolling mill, a motor (Marathon electric) and gearbox

(Baldor Electric Co.) are applied and connected by a coupler (Lovejoy) to the roller shaft as

seen in Fig. 3.1. Power is transmitted through the coupler to the bottom roller which makes

the bottom roller the drive roller. When specimens are inserted between the two rollers, the

top roller turns due to the friction between the specimens and the top roller; this makes the

top roller the driven roller.


12

Bottom roller

Coupler

Dynamometer
Motor

Fig. 3.1 Scheme for the motor and the bottom roller of the skew rolling setup

To adjust the rotation speed, a variable frequency drive (VFD) (Toshiba, VF-S15) is

employed. By connecting this VFD to the motor, the rotation speed of the rollers can be

controlled. Rotational speed adjustment is the only function that is required for the current

research, although this VFD offers a variety of functions which can be explored. Details of

the VFD can be seen in Fig. 3.2.


13

Fig. 3.2 Detailed view of the Toshiba VFD used to adjust the RPM

3.3. Design to change reduction ratios

The gap between the two rollers can be adjusted by various methods. Since skewness

is one of the main functions of this setup, common designs to adjust the gap cannot be

applied here. In our setup, slotted shims with various thicknesses were inserted underneath

the screws of the top roller. Various gaps of the two rollers can be achieved by alternating the

number of shims along with various thicknesses of shims, thus raising the height of top roller.
14

By changing the gap, the reduction ratios can be while the specimens’ thicknesses are held

constant. Fig. 3.3 shows the shape of shims used in the setup.

Fig. 3.3 Shims used in the setup to adjust gap between two rollers

3.4. Design to realize the skewness function of the setup

The most common rolling mills are parallel rolling mills, but in this study the top

roller can have an angle positioned up to ±15° oblique to the bottom roller. The power supply

for the top roller could be difficult given this fact. To achieve the goal of the skew roller, the

upper plates that support the top rollers at both ends are made with arc slots that create the

angles that can be seen in Fig. 3.4. The angle is adjusted by loosening the screw; moving the

pillow blocks along the arc slots to a desired angle; and retightening the screw.
15

Fig. 3.4 Solidworks drawing for the upper plate to achieve required angle

3.5. Design to acquire forces

3.5.1. Dynamometer

To acquire forces in the rolling process, a piezo-dynamometer (Kistler 9257B) has

been installed underneath the pillow block at one end, close to the motor of the bottom roller,

as can be seen from Fig. 3.1. The dynamometer is designed to have compactness, great

rigidity, and high natural frequency. The dynamometer setup is corrosion-resistant and is

protected against of spray water and cutting fluids. Forces of three axes can be measured by

this dynamometer with the specifications in Table 3.1.


16

Table 3.1 Specifications of the dynamometer employed in the setup


X axis Y axis Z axis

Measurement
-5-5 -5-5 -5-10
Range (kN)

Sensitivity (pC/N) 7.94 7.93 3.71

Sensitivities of the three axes are listed, and these will be applied to calibrate the

dynamometer. Ranges of the three axes were seen to be wide in the later experiments.

3.5.2. Data acquisition interface

An amplifier was employed to amplify the dynamometer’s signal. Physical data were

collected through the data acquisition card inserted in the computer. A LabVIEW program

was created to record data and for the researchers to operate. Fig. 3.5 shows the force

measurement system in this setup.

Dynamomet LabVIEW
Amplifier Data acquisition card
er interface

Fig. 3.5 Force measurement setup flowchart in the rolling mill

Fig. 3.6 and Fig. 3.7 present the LabVIEW interfaces that are designed to record

forces.
17

Fig. 3.6 LabVIEW interface front panel. The white, red, greed curves stand for X, Y, and Z
forces

Fig. 3.7 Block diagram behind front panel of the LabVIEW program
18

The white, red, and green curves in the three windows from top to bottom in Fig. 3.6

are the real-time forces of the X, Y, and Z axes. The curves can be exported to other files for

further investigation. Fig. 3.7 shows the block diagram behind the front panel to portray how

the physical signals are converted to forces. Signals from the data acquisition card are read

by this program and are converted to forces based on the settings of the amplifier. Three

channels of force data are illustrated by waveform charts.

3.5.3. Dynamometer calibration

To obtain accurate force data, the dynamometer needs to be calibrated. The common

method for calibration is to use standard weights in three axes. The standard weights are

compared with the readings from the dynamometer to find the adjustment factor. Due to

space limitation, two axes in this rolling setup could not be accomplished. The amplifier in

this system, however, offers a second way to calibrate.

Fig. 3.8 Dynamometer shows the sensitivities and amplification factors of three axes
19

As shown in Fig. 3.5, the dynamometer is connected to an amplifier that is employed

to enhance the ability to acquire data. To calibrate the dynamometer, the sensitivities of three

axes from Table 3.1 are applied in the first row of amplifier. To meet various ranges of the

three axes, amplification factors in three axes are adjusted both in the amplifier and in the

block diagram. Values of amplification factors in the amplifier should remain consistent with

values in the equations in order to calculate forces in three axes in the block diagram. Details

can be seen in Fig. 3.8. As the figure shows, the three values of the first row are the

sensitivities of the X, Y, and Z axes. The second row of the amplifier is the amplification

factors of three axes. Values are consistent from Table 3.1, Fig. 3.7, and Fig. 3.8. After the

calibration, standard weights were placed on the dynamometer to validate the whole

procedure and the results showed that the calibration was effective and successful.

CHAPTER 4. FORCE ANALYSIS AND GEOMETRY MODELS IN THE ROLLING

PROCESS

4.1. Force analysis and geometry model of parallel rolling

Rolling is the process of reducing the initial thickness by compressive forces that are

applied by rolls. Parallel rolling is a process in which the axes of two rollers are parallel to

each other. Fig. 4.1 shows the schematic deformation of the specimens in between the two

rollers. h0 is the thickness of the specimen. hf stands for the final thickness, which is equal to

the gap between two rollers. w is the width of the specimen. R is the radius of rollers. The

two rollers in our setup have the same radius.


20

Fig. 4.1 Deformation scheme in parallel rolling process

Fig. 4.2 Schematic drawing of Sp

The reduction is relatively small compared to the radius of the roller; the deformed

area Sp in XY plane seen in Fig. 4.2 can be calculated as:

S p  w Rh0  h f  (1)
21

The force in Z axis Fzp that is due to the reduction can be easily calculated as the

product of the pressure P in the Z axis and the deformed area in the XY plane.

Fzp  P  S p (2)

The force in the Y axis Fyp can be represented as the product of the coefficient of

friction μ and the force in the Z axis.

Fyp    Fzp (3)

In the parallel rolling process, there is no component force in the X axis.

Fig. 4.3 shows the force analysis for the bottom roller. For simplicity, the pressure

distribution is assumed to be uniform along the width of the roller. The dynamometer is

placed underneath one of the two ends of the bottom roller as seen from Fig. 3.1.

Fig. 4.3 Forces analysis on bottom roller in parallel rolling process


22

Fzd and Fyd are the forces measured by the dynamometer, and they are shown in the

LabVIEW program. From the forces scheme on the bottom roller shown in Fig. 4.3, it would

be straightforward to find the equation ( 4 ) ( 5 ) ( 6 ) ( 7 ) due to the symmetry:

1
Fzd  Fzp (4)
2

Pw Rh0  h f 
1
Fzd  (5)
2

1
Fyd  Fyp (6)
2

Pw Rh0  h f 
1
Fyd  (7)
2

In a later study, the forces shown on the LabVIEW program are compared with the

forces calculated following these relations above.

4.2. Force analysis and geometry model of skew rolling

The top roller can be rotated to have an angle  ranging from -15° to 15° obliquely to

the bottom roller through the angle guidance. A specimen is positioned and rolled as shown

in Fig. 4.4. x is the contact width of which the value should be no greater than the specimen

width w.
23

Fig. 4.4 Specimen in between the two rollers, the red rectangle is the maximum area it can
deform, the arc shaped area is an approximate area assumed

From the Fig. 4.4, it is simple to determine the maximum deformed area, which is

shown as a rectangle Ss:

 
S s  tan  x 2 (8)
2

The edges of the red rectangles would not be in contact with either roller in the actual

experiments. The actual contact area would be more accurately represented as an arc shaped

area. Assume the curves are parts of circles; and then the arc shaped area would be:

   
tan 2   x 2 (   ) x 2 tan 3  
  x2 2
S a  tan  x 2  
x
 2  2
(9)
2       2
4 sin 2   2 tan  4 cos 2  
2 2 2

The area ratio Ra is defined as:

Ss
Ra  (10 )
Sa
24

Due to various materials properties, deformation would be different when different

materials were used. Consequently, deformed areas would be different even with the same

reduction ratio. An area factor would have to be introduced to calculate the real deformation

area.

The contact curve from the center point to the edge of the specimen between the

bottom roller and the lower surface of the specimen can be simplified as seen in Fig. 4.5:

Fig. 4.5 Specimen in contact with the bottom roller surface. Contact curve is represented as
the ellipse dot line
25

Points on the ellipse curve should fall into the equation ( 11 ):

2
 x 
 
 2 cos   
 2     R  h  h f
2
 
     1 ( 11 )
R
   R 
 sin   
 
 

h is the distance from any point on the curve to the bottom line of the top roller, as can be

seen in Fig. 4.5.

To calculate the area Ss, x needs to be determined. Firstly, let x equal w, the specimen

width, to obtain the maximum height hmax. Compare the specimen thickness h0 with hmax. If

h0 is smaller than hmax , it indicates that part of the specimen is not in contact with the bottom

roller’s surface. The value x can be obtained by solving the equation ( 11 ), under h= h0. If h0

is larger than hmax, it indicates that even the edge of the specimen is deformed. The contact

width is w while the thickness of the specimen at the edge is not h0, but hmax.

Due to the skewness of the rollers, there exists a component force in the X axis in the

bottom roller; and this is a distinct difference compared to parallel rolling.


26

Fig. 4.6 Force analysis on bottom roller under skew rolling process

As shown in Fig. 4.6, Fzs is the pressure force on bottom roller that is due to the

reduction. Fzs can be simply represented by the product of the pressure P in the Z axis and the

deformed area in XY plane.

F zs P  S a ( 12 )

Fys is the component friction force caused by the pressure.

 
Fys  Fzs cos  ( 13 )
2

Fxs is also the component friction force caused by the pressure.


27

 
Fxs  Fzs sin  ( 14 )
2

Fzd and Fyd are the forces measured by the dynamometer which are shown in the

LabVIEW program. From the forces scheme on the bottom roller shown in Fig. 4.3, it would

be straightforward to find the equation ( 15 ) ( 17 ) due to the symmetry:

1
Fzd  Fzs ( 15 )
2

1   x 2 x2
Fzd  P[tan  x 2  
2 2    
4 sin 2   2 tan 
2 2

( 16 )
   
tan   x 2 (   ) x 2 tan 3  
2

 2   2 ]
  2
4 cos 2  
2

1
Fyd  Fys ( 17 )
2

1     x 2 x2
Fyd  P cos [tan  x 2  
2 2 2    
4 sin 2   2 tan 
2 2

( 18 )
   
tan   x 2 (   ) x 2 tan 3  
2

 2   2 ]

  2
4 cos 2  
2

However, due to the over constraint of bottom roller in the X axis direction, the

relationship between Fxd and Fxs is not determined.


28

4.3. Experimental validation of force analysis

4.3.1. Experimental data, calculations and simulations of parallel rolling

To validate the analyzed forces, some experiments were conducted. Two materials,

PETG (polyethylene terephthalate-glycol) and HDPE (high-density polyethylene) were

chosen to be rolled with different reductions. A finite element analysis (FEA) simulation was

also performed to obtain the resultant forces. The drawing in Abaqus was based on the

original Solidworks files for fabrication. The top roller and bottom roller were set as analytic

rigid parts in order to save simulation time. For later experiments, only plastics were inserted

in between the two rollers. Because of this, neglecting the deformation of the rollers would

not introduce significant error.

Fig. 4.7 Rolling model created in Abaqus


29

The position of rollers was locked. All displacements and rotations were set to zero

except the rotation about the X axis in Fig. 4.7. The specimen was initially positioned in

between the two rollers and an angular velocity of the bottom roller was introduced when the

specimen started being rolled. Properties of the specimens followed the two curves in Fig.

4.10 and Fig. 4.11. Resultant forces were obtained after the rolling.

Table 4.1 Experimental settings of parallel rolling for HDPE and PETG
Final Thickness, Reduction, Width, w Average Roller Coefficient

thickness, h0 (mm) h0-hf, (mm) pressure, radius, R of friction, µ

hf (mm) (mm) P (MPa) (mm)

HDPE-1 1.0922 1.5241 0.4319 50.8 23.4 69.85 0.1

HDPE-2 2.286 3.048 0.762 50.8 23.4 69.85 0.1

PETG-1 1.27 1.397 0.127 50.8 52 69.85 0.05

PETG-2 2.54 2.794 0.254 50.8 52 69.85 0.05

Fig. 4.8 and Fig. 4.9 show the forces measured by the dynamometer. Due to the

unbalanced weight of the roller itself, there are some fluctuations when there were no

specimens in between the rollers. Compared to the measured values, the error caused by the

imbalance can either be deducted from the total values or it can be neglected due to its small

magnitude.

The curves in Fig. 4.8 and Fig. 4.9 are not flat when the specimen is being rolled and

measured. One reason might be small vibrations since the rollers are not perfectly rigid; the

second reason might be the specimens didn’t move straight which made the deformed area
30

varying in the rolling process. The average values of the curves when the specimens were

being rolled were recorded on the table.

700

600

500
HDPE-1
400
Fyd (N)

HDPE-2
300
PETG-1
200
PETG-2
100

0
0 500 1000 1500 2000
-100
Time

Fig. 4.8 Measured Y axis force under parallel rolling with different reductions

6900

5900

4900
Fzd (N)

3900
HDPE-1
2900 HDPE-2

PETG-1
1900
PETG-2
900

-100
0 500 1000 1500 2000
Time

Fig. 4.9 Measured Z axis force under parallel rolling with different reductions
31

Forces of HDPE-2 and PETG-2 were much larger than forces of HDPE-1 and PETG-

1 both in the Z axis and the Y axis directions. From equation ( 3 ), it can be seen that the Y

axis force is proportional to the Z axis forces. The thickness of HDPE-2 and PETG-2 was

reduced more, leading to more deformation and more force. Fig. 4.8 and Fig. 4.9 show the

forces in the Z axis have increased approximately by 2000 N for both materials. Due to

different frictional conditions for two materials, the forces increase in the Y axis direction by

different amounts.

To calculate the force by applying equations ( 5 ) and ( 7 ), two more parameters need

to be fully understood: The average pressure and the coefficient of friction. The coefficient of

friction varies greatly with different conditions. It would not be accurate enough to find the

values from literature since it would be difficult to find exactly the same conditions in other

operations. Due to this limitation, it’s reasonable to assign values and validate the assigned

values using other conditions. In our setup, the coefficients of friction for PETG and HDPE

in contact with the low carbon steel roller are assumed to be 0.05 and 0.1 respectively.

The rolling is a process under plane stress. Compressive stress strain curves were

found in the literature to offer the deformation information for the current experiments. The

following two figures Fig. 4.10 and Fig. 4.11 from Moura et al. [34], as well as Dupaix and

Boyce [35] are applied to show the compressive stress strain relation of HDPE and PETG.
32

Fig. 4.10 Compressive stress strain curve for HDPE, green curve in the graph is applied

Fig. 4.11 Compressive stress strain curve for PETG, 0.5/s plane in the graph is applied

The average pressure is obtained by calculating the strains of the deformed specimen

and the corresponding stress from those curves. The average pressure is the average Y axis

value in Fig. 4.10 and Fig. 4.11. All of the pressures are listed on the table. After the forces

equations in the parallel rolling are determined, Fyp and Fzp can be calculated.
33

Table 4.2 Comparison between the measured forces and the calculated forces in parallel
rolling
Experiment data Calculation data Simulation data

Fyd (N) Error (N) Fzd (N) Error (N) Fyd (N) Fzd (N) Fyd (N) Fzd (N)

HDPE-1 320 29.7 3815 330.2 326 3264 291 3778

HDPE-2 500 71.2 5600 787.4 445 4450 356 5816

PETG-1 175 29.2 3900 330.2 197 3934 222 3858

PETG-2 281 42.5 6238 516.7 288 5563 356 6489

The calculated values are plotted in Fig. 4.12 and Fig. 4.13. As can be seen, the

calculated forces are close to the measured values either in the Y axis or the Z axis.

Experimental and simulation data are close to each other in the Z axis forces in Fig. 4.13.

However, calculation data are generally smaller. The calculation assumes the average

pressure in the deformed area, which may result in smaller values. For the comparison in Fig.

4.12, the simulation is smaller than that of the experiment data of HDPE-2 while the

calculation falls into the range of the experimental data. The simulation of HDPE-2 is smaller

than the experiment data and this can be explained by the fact that the theoretical friction

condition in the simulation may not accurately represent the real experiment. The friction in

actual situation for HDPE might be higher than an assumed single value in the simulation.

For the PETG in Fig. 4.12, the experimental data and calculation data are close, while the

simulation data is larger. Deflection is the main reason behind this phenomenon. Since the

rollers are considered to be rigid in the simulation, simulation data are usually larger than

experimental data. Since the PETG is harder than HDPE, the deflection would be more

evident for PETG rather than for HDPE. As a result, the friction assumed in the simulation
34

factor dominated the difference rather than the rigidity of HDPE. Consequently, the

difference between the simulation and experiment of PETG would be more distinct than the

difference between the simulation and experiment for HDPE. In the figure, it can be seen the

Z axis force of PETG-2 is higher than the experimental force, which leads to a higher force in

the Y axis.

Overall, the results shown in the figures can lead to the conclusion that the assumed

coefficient of friction can correctly represent the actual friction conditions in the rolling setup,

since different reductions were carried out for two materials in which experimental data,

calculated data, and simulated data precisely met each other.

The validation of the coefficients of friction for two different materials in turn proved

the correctness of the equations and the assumptions behind those equations.

600

500
Force in Y axis (N)

400

Fyd Measured
300
Fyd Calculated

200 Fyd Simulation

100

0
HDPE-1 HDPE-2 PETG-1 PETG-2

Fig. 4.12 Comparison among the measured forces, the calculated forces and simulated forces
in Y axis in parallel rolling
35

7000

6000

5000
Force in Z axis (N)
4000
Fzd Measured
3000
Fzd Calculated
2000 Fzd Simulation

1000

0
HDPE-1 HDPE-2 PETG-1 PETG-2

Fig. 4.13 Comparison among the measured forces, the calculated forces and simulated forces
in Z axis in parallel rolling

4.3.2. Experimental data, calculations and simulations of skew rolling process

After the validation of the parallel rolling, the coefficients of friction of both materials

are demonstrated to be accurate. In the skew rolling, the coefficients of friction are assumed

to be the same and known.

In the skew rolling, experiments were conducted under three different angles between

two rollers: 15°, 10°, and 6°. Other parameter settings can be seen from Table 4.3.
36

Table 4.3 Experiment setting of skew rolling process with PETG and HDPE

Trials PETG-15° PETG-10° PETG-6° HDPE-15° HDPE-10° HDPE-6°

Final thickness, hf
1.143 1.143 1.143 1.143 1.143 1.143
(mm)
Thickness at the
1.397 1.2827 1.1938 1.4968 1.2827 1.1938
edge, h (mm)
Contact width, x
42.672 50.8 50.8 50.8 50.8 50.8
(mm)
Average pressure,
58 66 62 22 28 28
P (MPa)
Roller radius, R
69.85 69.85 69.85 69.85 69.85 69.85
(mm)
Coefficient of
0.05 0.05 0.05 0.1 0.1 0.1
friction, µ
Maximum
deformed area, Ss 239.73 225.78 135.25 339.75 225.78 135.25
(mm2)
Arc shaped area,
139.08 136.73 84.98 197.07 136.73 84.98
Sa (mm2)

Ratio, Ra 1.72 1.65 1.59 1.72 1.65 1.59

Area factor 1.3 1.3 1.3 1.55 1.55 1.55

Fig. 4.14, Fig. 4.15, Fig. 4.16 and Fig. 4.17 show the measured results of the

dynamometer in the Y axis and the Z axis.

510
PETG Y axis Force (N)

410

310
PETG 15°
210 PETG 10°

110 PETG 6°

10
0 500 1000 1500 2000
-90
Time

Fig. 4.14 Measured force in Y axis of PETG under different angles


37

600

500

400
HDPE Y axis Force (N)

300
HDPE 15°
HDPE 10°
200
HDPE 6°

100

0
0 500 1000 1500 2000 2500

-100
Time

Fig. 4.15 Measured force in Y axis of HDPE under different angles

5100

4100
PETG Z axis force (N)

3100

PETG
2100 15°
PETG
10°
1100

100
0 500 1000 1500 2000
-900
Time

Fig. 4.16 Measured force in Z axis of PETG under different angles


38

6000

HDPE Z axis force (N) 5000

4000

3000
HDPE 15°
2000 HDPE 10°
HDPE 6°
1000

0
0 200 400 600 800 1000 1200 1400
-1000
Time

Fig. 4.17 Measured force in Z axis of HDPE under different angles

Forces in skew rolling are different when compared with the forces in parallel rolling,

because there would have a component force which would cause a slight slip in the X axis.

As a result, the deformation area would change with the motion. This is the main reason why

the curves in the figures are not flat when the specimens were being rolled.

The motion of the specimen under skew rolling contributes to the inclination of the

curves in the Z axis. The specimen would move away from the center and, as a result, the

deformation area would be smaller. The reduction of the area would result in a smaller force,

as can be seen from the equation: The force is the product of the pressure and the deformed

area.

Table 4.4 shows the calculated forces in comparison to the forces measured.
39

Table 4.4 Comparison between the measured forces and the calculated Y axis and Z axis
forces for PETG and HDPE at three different angles
Experiment data Calculation data

Fyd (N) Error (N) Fzd (N) Error (N) Fyd (N) Fzd (N)

PETG-15° 320 55.6 4600 351.1 260 5254

PETG-10° 300 26.0 4500 497.2 292 5870

PETG-6° 190 28.3 3500 243.4 172 3428

HDPE-15° 380 64.4 3600 211.2 333 3368

HDPE-10° 330 39.7 3500 278.6 296 2969

HDPE-6° 270 38.9 2500 404.7 185 1846

Table 4.4 is plotted based in Fig. 4.18 and Fig. 4.19.

450
400
350
Force in Y axis (N)

300
250
Fyd Measured
200
Fyd Calculated
150
100
50
0
PETG-15° PETG-10° PETG-6° HDPE-15° HDPE-10° HDPE-6°

Fig. 4.18 Comparison between the measured forces and the calculated forces in Y axis in
skew rolling for PETG and HDPE at three different angles
40

6000

5000

Force in Z axis (N) 4000

3000 Fzd Measured

Fzd Calculated
2000

1000

0
PETG-15° PETG-10° PETG-6° HDPE-15° HDPE-10° HDPE-6°

Fig. 4.19 Comparison between the measured forces and the calculated forces in Z axis in
skew rolling for PETG and HDPE at three different angles

As can be seen in Fig. 4.18 and Fig. 4.19, the calculated Z axis forces of HDPE were

smaller than the experimental forces. Because HDPE is more easily deformable than PETG,

under the skew rolling, the contact area would be larger than the contact area of PETG. The

difference between the experimental observations with the calculation of HDPE could be the

area consideration. The contact area in the actual situation could be larger than assumed.

Since the Y axis force is proportional to the Z axis force, the Y axis force of the HDPE is

also smaller than the experimental force.

For the comparison of PETG, the contact area under a skew angle of 10 would be

larger than the contact area under a skew angle of 15 based on the geometry of the model.

From the point of the geometry model, part of the specimen was not rolled under a skew

angle of 15. From the equations used to calculate forces, forces under a skew angle of 10

would be larger than forces under a skew angle of 15. Evidently, the experimental
41

observations did not follow the trend. However, the experimental data under skew angles of

10 and 15 are close. One explanation could be that the slip in the X axis caused a change in

the contact area and this variance was not considered in the equations. The experimental Z

axis forces are smaller than those calculated. This could be due to the fact that the PETG is

harder than HDPE. With the same reduction, PETG would be harder to deform. The

deformed area was smaller than expected. Also in the skew rolling process, the specimens

were supposed to be fed through in the centerline between two rollers. As the specimen

might not be fed in accurately along the centerline of two rollers, the specimen might have a

greater component force in the Y axis than expected. That contributed to larger experimental

data than calculations.

Overall, Fig. 4.18 and Fig. 4.19 successfully demonstrate the force models under

skew rolling. Due to the setup’s rigidity, there would be deflections in the rolling process that

could contribute to the errors in the attempts to predict the exact values. The forces would

drive the specimen away from the center, but the area calculation in the model is based on the

contact center of the two rollers. This difference will also cause some errors compared to the

actual data.
42

CHAPTER 5. EXPERIMENTAL STUDY ON THE ANGLE COMPENSATION

In the skew rolling process, the component force in the axial direction would be

generated due to the crossed angle. This component force is the force that pushes the

bearings in the rolling setup. The influence of this force on the bearings needs to be

understood. The influence of rolling process parameter, the crossed angle, and the friction

conditions, on the force ratio Fxd/ Fzd was analyzed.

5.1. Experimental procedure and results

Two specimens, PETG and HDPE, were rolled under eight different angles. Forces in

the X axis Fxd and the Z axis Fzd, were measured. Simulation has been performed under the

same conditions. Measured data were compared to simulated data and discussed in the

following context.

Table 5.1 is the experimental setting for the two specimens.

Table 5.1 Experiment setting of two specimens


Final Thickness, Width, Roller Coefficient

thickness, h0 (mm) w (mm) radius, of friction,

hf (mm) R (mm) µ

HDPE 1.016 1.4732 50.8 69.85 0.1

PETG 1.016 1.397 50.8 69.85 0.05


43

Fig. 5.1 to Fig. 5.4 are some examples of forces at various angles. Results of other

angles are listed in the table.

250

200
X axis force of HDPE (N)

150

100

50

0
8000 8100 8200 8300 8400 8500 8600

-50

Fig. 5.1 X axis force of HDPE at an angle of 6

4500

4000
Z axis force of HDPE (N)

3500

3000

2500

2000 3°

1500

1000

500

0
6000 6200 6400 6600 6800 7000
-500

Fig. 5.2 Z axis force of HDPE at an angle of 3


44

300

250
X axis force of PETG (N)

200

150

100

50

0
1200 1250 1300 1350 1400 1450 1500

-50

Fig. 5.3 X axis force of PETG at an angle of 3

5800

4800
Z axis force of PETG (N)

3800

2800 8°

1800

800

-200
13900 13950 14000 14050 14100 14150 14200 14250 14300 14350 14400

Fig. 5.4 Z axis force of PETG at an angle of 8


45

Forces in the Z axis were generally flatter at different angles. One reason was that the

error of the unbalanced weight of the rollers was relatively small compared with the rolling

forces. Another reason might be that forces in the XY plane could cause the vibration, with

the vibration contributing to the difficulty of recording forces in the X or the Y axis.

Table 5.2 Experiment X and Z axes forces of HDPE and PETG at eight different angles
HDPE PETG

Fxd (N) Error (N) Fzd (N) Error (N) Fxd (N) Error (N) Fzd (N) Error (N)

1
139 24 3450 350 135 25 3479 431
2
210 38 3561 389 195 41 3970 580
3
202 26 3522 298 195 23 3800 370
4
234 16 3705 460 200 23 3822 343
5
241 19 3653 197 225 45 4033 367
6
220 25 3700 200 230 39 4200 370
8
229 33 3772 194 255 55 4460 340
10
197 22 3811 214 225 64 3884 816

Fig. 5.5, Fig. 5.6, Fig. 5.7 and Fig. 5.8 are the plotted based on Table 5.2.
46

300

250

X axis force of HDPE (N)


200

150

100

50

0
1° 2° 3° 4° 5° 6° 8° 10°

Fig. 5.5 X axis forces of HDPE at different angles

4500

4000

3500
Z axis force of HDPE (N)

3000

2500

2000

1500

1000

500

0
1° 2° 3° 4° 5° 6° 8° 10°

Fig. 5.6 Z axis forces of HDPE at different angles


47

350

300

X axis force of PETG (N)


250

200

150

100

50

0
1° 2° 3° 4° 5° 6° 8° 10°

Fig. 5.7 X axis forces of PETG at different angles

5000
4500
4000
Z axis force of PETG (N)

3500
3000
2500
2000
1500
1000
500
0
1° 2° 3° 4° 5° 6° 8° 10°

Fig. 5.8 Z axis forces of PETG at different angles

Although the forces did not change too much, it could still be found out that rolling

forces increase slowly as the angle increases. From equation ( 9 ), an increasing angle would

lead to a larger contact area. With the force as a product of area and stress, rolling forces
48

were expected to increase. Forces in the axial direction were not only affected by the rolling

forces, but were also affected by the crossed angle. Component force Fxd in the axial

direction would decrease with the increasing angle, but the increased angle would have a

larger rolling force that would increase the component force. From Fig. 5.5 and Fig. 5.7,

forces in the axial direction increased from the beginning but decreased at larger angles.

0.18

0.16

0.14

0.12
Simulated Fx/Fz
Force Ratio

0.1 with coefficient of


friction 0.2
0.08 Simulated Fxd/Fzd
with coefficient of
0.06 friction 0.1
0.04 Experimental
Fxd/Fzd
0.02

0
0 2 4 6 8 10 12
Angles ()

Fig. 5.9 Comparison between simulation and experiment data of Fxd/ Fzd of HDPE
49

0.2
0.18
0.16
0.14
Simulated Fx/Fz with
Force Ratio

0.12 coefficient of friction


0.1 0.1
Simulated Fxd/Fzd
0.08 with coefficient of
0.06 friction 0.05
Experimental Fxd/Fzd
0.04
0.02
0
0 2 4 6 8 10 12
Angles ()

Fig. 5.10 Comparison between simulation and experiment data of Fxd/ Fzd of PETG

5.2. Discussion

From the results in section 5.1, the force ratio increased with the angle significantly

from zero to around 0.06 for HDPE and 0.05 for PETG. It can be seen in Fig. 5.9 and Fig.

5.10 that the blue curve and the green curve, which represented the simulation and

experimental data respectively, plateaued after 2 for HDPE and 6 for PETG. In Fig. 5.9,

both the blue curve and green curve showed that an angle of 2 was the threshold for HDPE

specimen.

This indicates the force ratio in the rolling process of HDPE was sensitive in the

range 0-2 and that it would increase significantly with a slight angle change. Beyond the

threshold angle of 2, the force ratio would be less sensitive to the angle which was reflected

from the figure as an almost flat curve with increasing angles. In Fig. 5.10, both simulation

and experimental data showed a similar trend for the PETG specimen. Although the
50

experimental curve was off by the simulation curve 0.02, they shared the same trend, and it

could be seen from the curve that the ratio increased significantly from 0 to 2, and then

slightly increased from 2 to 6. After 6, the curve became flat. The experimental curve was

less sensitive as the angle changed from 2 to 6 compared to the simulation curve. The

deflection of the rollers in experiment processing should be taken into consideration to

explain the slight variations between the simulation and the experiments. Other factors

include the position in which specimens were inserted, the variation of the thickness of

specimens, the gap change due to deflection, and so forth.

Simulation data with different frictional conditions were also plotted as the red curves

in both specimens. As can be seen from the Fig. 5.9 and Fig. 5.10, if the coefficient of

friction doubled, the force ratio would be 100% larger, which was expected since the axial

force would increase due to the coefficient of friction. The red curves didn’t show the

threshold angles, but did show a continuous increase.

The axial force in the rolling system is the force used to push the bearings, it would

be meaningful to understand and control the variations. The study carried out in this chapter

showed that for a lower coefficient of friction, the force ratio varied little after 2 for HDPE

and 8 for PETG. The force ratio can be minimized by compensating for this threshold angle

instead of for a larger one because they have the same force ratio. As shown in Fig. 5.9, the

force ratio under an angle of 2 had the same value as with an angle of 10. To minimize the

axial force under 10, only the 2 needed to be compensated. In Fig. 5.10, the threshold

angle would be 6. This trend which could be used to the experimenter’s advantage

disappeared when there were larger coefficients of friction. That means that the exact angle

needs to be compensated for, in order to minimize the axial forces. Trend difference between
51

larger coefficients of friction and lower coefficients of friction indicated that lubricants

would benefit the compensation process by minimizing the force ratios.


52

CHAPTER 6. CONCLUSIONS

1. Hardware design and fabrication of the rolling mill provided the functions needed

to make the gap as small as 0.25 mm; the roller can be rotated at speeds ranging from 0 to 30

RPM; a dynamometer was successfully calibrated and could accurately measure the forces in

the ranges of three axes; angles could be adjusted as accurately as 1, and the range could be

up to 15.

2. Deformation area models for parallel rolling and skew rolling have been built.

Experiment data were measured and compared to the calculated values. Coefficients of

friction for PETG and HDPE materials in this rolling setup were determined to be 0.05 and

0.1 respectively. Calculations and experimental data showed that the rolling force can be

calculated as a product of stress and the deformed area. The thrust force can be calculated as

a product of rolling force and the coefficient of friction. Based on the friction conditions

determined in the former section, equations in skew rolling were shown to be validated.

Calculations and experimental data showed that the rolling force can be calculated as a

product of stress and the deformed area. The deformed area followed the geometry built in

the former chapter. Forces in other two axes can be calculated as two component forces of a

product of rolling force and coefficient of friction.

3. Throughout this study, the influence of rolling process parameter, the crossed angle,

and the friction conditions on the force ratio Fxd/ Fzd was analyzed. The study carried out

showed that for a lower coefficient of friction, the force ratio varied little after 2 for high

density polyethelene (HDPE) and 6 for polyethylene terephthalate glycol-modified (PETG).

The force ratio can be minimized by compensating for these threshold angles instead of
53

actual larger angles because they have the same force ratio. Force ratios with larger

coefficients of friction losing this similar trend suggest that attempts to lower the coefficient

of friction were necessary. As the axial force is the force that pushes the bearings in the setup,

minimizing this force is necessary. The trend in Fig. 5.9 and Fig. 5.10 can be obtained with a

lower coefficient of friction, and angle compensation can be carried out based on this trend.

Analysis of the force ratio provided methods to minimize the forces, which is through

lubricants and angle compensation. This study can also be expanded to other materials under

different friction conditions to obtain threshold angles for control purposes.


54

CHAPTER 7. BIBLIOGRAPHY

[1] Nishioka, K., Mizutani, Y., Hori, Y., Kojima, A., and Ogawa, S., 1997, "High Reduction
Rolling Technology on Pair Cross Mill," Nippon Steel Technical Report, 75(0), pp.

[2] Omori, S., Hino, H., and Goto, H., 1994, "Characteristic Phenomena Observed During
Rolling by Pair Crossed Mill," Ironmaking & steelmaking, 21(6), pp. 465-471.

[3] Omori, S., Kajiwara, T., Hino, H., and Goto, H., 1997, "Walking Characteristic of Pair
Crossed Rolling Mill," Ironmaking & steelmaking, 24(4), pp. 337-344.

[4] Ginzburg, V. B., and Azzam, M., 1996, "Selection of Optimum Strip Profile and Flatness
Technology for Rolling Mills," eds., pp. 215-232.

[5] Nishibayashi, S., Hatae, S., Haga, M., and Hirase, K., 1986, "Design and Operation of a
New Hot Strip Mill at Hirohata Works," Iron Steel Eng., 63(10), pp. 49-56.

[6] Shimizu, Y., Iwatani, J., and Hasai, J., 1999, "Compact Hot Strip Mill for High Quality
Strip Production," Mitsubishi Juko Giho(Japan), 36(6), pp. 296-299.

[7] Omori, S., Kajiwara, T., Hino, H., and Goto, H., 2004, "Analysis of Rolling Load
Generated by Pair Crossed Rolling Mill," Ironmaking & steelmaking, 31(1), pp. 71-80.

[8] Nakasuji, K., 1995, "Effects of Rolling Conditions on Rolling Load, Rolling Torque and
Power Consumption in Heavy Reduction Rolling by the Rotary Reduction Mill," ISIJ
International, 35(2), pp. 188-196.

[9] Aljabri, A., Jiang, Z. Y., Wei, D. B., Wang, X. D., and Tibar, H., 2014, "Thin Strip
Profile Control Capability of Roll Crossing and Shifting in Cold Rolling Mill," eds., 773, pp.
70-78.

[10] Qiu, J. H., Murata, T., Takahashi, K., and Wu, X. L., 2012, "The Plastic Deformation
Characteristics of Crystalline Polymer Materials Via Rolling Process," Advanced Materials
Research, 391(pp. 585-589.

[11] Vihari, K., and Kumar, S., 1980, "Cold Ring Rolling of a Polymer," International
Journal of Machine Tool Design and Research, 20(2), pp. 97-109.

[12] Murata, T., Qiu, J. H., Sakai, E., Wu, X. L., and Kudo, M., 2012, "Morphology Change
of Extrusion Molded Crystalline Polymer by Rolling Process," Advanced Materials Research,
391(pp. 595-599.

[13] Dieter, G. E., and Bacon, D., 1986, Mechanical Metallurgy, New York: Mcgraw-Hill,
55

[14] Degarmo, E. P., Black, J. T., and Kohser, R. A., 2011, Degarmo's Materials and
Processes in Manufacturing, John Wiley & Sons,

[15] Kalpakjian, S., and Schmid, S. R., 2010, Manufacturing Processes for Engineering
Materials, Pearson education,

[16] Dhingra, V. J., Spruiell, J. E., and Clark, E. S., 1981, "The Relationship between
Mechanical Properties and Structure in Rolled Polypropylene," Polymer Engineering &
Science, 21(16), pp. 1063-1071.

[17] Cangialosi, D., Wübbenhorst, M., Schut, H., Van Veen, A., and Picken, S., 2005,
"Amorphous-Amorphous Transition in Glassy Polymers Subjected to Cold Rolling Studied
by Means of Positron Annihilation Lifetime Spectroscopy," The Journal of chemical physics,
122(6), pp. 064702.

[18] Matsuoka, S.-I., 1998, "Effects of Cold-Rolling on the Cold Forging of Solid Polymer,"
Journal of Materials Processing Technology, 84(1–3), pp. 175-180.

[19] Kitagawa, M., Isobe, S., and Asano, H., 1982, "Crack Growth under Static and Fatigue
Loading in Glassy Polymers Oriented by Cold-Rolling," Polymer, 23(12), pp. 1830-1834.

[20] Asano, T., and Fujiwara, Y., 1978, "Plastic Deformation of Oriented Lamellae: 1. Cold
Rolling of Β-Phase Isotactic Polypropylene," Polymer, 19(1), pp. 99-108.

[21] Caddell, R. M., Bates Jr, T., and S.Y. Yeh, G., 1972, "On the Tensile Behavior of High
Density Polyethylene Subjected to Cold Rolling," Materials Science and Engineering, 9(0),
pp. 223-229.

[22] Shonaike, G. O., and Reed, P. E., 1989, "The Effect of Cold Rolling on Crazing of
Polycarbonate," Materials Science and Engineering: A, 119(0), pp. 231-237.

[23] Wang, B., 1996, "A Simulation of Roll Wear in Hot Rolling Processes," pp.

[24] Johnson, K. L., and Johnson, K. L., 1987, Contact Mechanics, Cambridge university
press,

[25] Hertz, H., 1881, "On the Contact of Elastic Solids," J. reine angew. Math, 92(156-171),
pp. 110.

[26] Carter, F., 1926, "On the Action of a Locomotive Driving Wheel," Proceedings of the
Royal Society of London. Series A, 112(760), pp. 151-157.

[27] Poritsky, H., 1950, "Stresses and Deflections of Cylindrical Bodies in Contact with
Application to Contact of Gears and of Locomotive Wheels," Journal of Applied Mechanics-
Transactions of the ASME, 17(2), pp. 191-201.
56

[28] Arnell, R. D., Davies, P., Halling, J., and Whomes, T., 1991, "Tribology: Principles and
Design Applications," New York, Springer-Verlag, 1991, 266 p., 1(pp.

[29] He, M., and Sawamiphakdi, K., 2004, "Application of Computer Programs to Assist the
Tooling Design for the Skew Rolling Process," Materials Science & Technology: September
26-29, 2004, New Orleans, Louisiana, pp. 113.

[30] Hu, Z., Xu, X., and Sha, D., 1985, "Skew Rolling and Cross Wedge Rolling—Principles,
Processes and Machines," Metall. Ind. Press, Beijing, China, pp.

[31] Yang, H. B., Zhang, L. J., and Hu, Z. H., 2012, "The Analysis of the Stress and Strain in
Skew Rolling," Advanced Materials Research, 538(pp. 1650-1653.

[32] Yang, S.-C., and Chen, C. O.-K., 2001, "The Surface Geometry of Rollers with Skew
Rolling of Steel Balls," Proceedings of the Institution of Mechanical Engineers, Part C:
Journal of Mechanical Engineering Science, 215(5), pp. 523-532.

[33] Hu, J. H., Shuang, Y. H., and Liu, P., 2011, "A Hybrid Model Using Genetic Algorithm
and Neural Network for Optimizing Technology Parameters in Skew Rolling Seamless
Pipes," Advanced Materials Research, 145(pp. 117-122.

[34] Moura, R. T., Clausen, A. H., Fagerholt, E., Alves, M., and Langseth, M., 2010, "Impact
on Hdpe and Pvc Plates – Experimental Tests and Numerical Simulations," International
Journal of Impact Engineering, 37(6), pp. 580-598.

[35] Dupaix, R. B., and Boyce, M. C., 2005, "Finite Strain Behavior of Poly(Ethylene
Terephthalate) (Pet) and Poly(Ethylene Terephthalate)-Glycol (Petg)," Polymer, 46(13), pp.
4827-4838.